Artificial intelligence (AI) is set to play a key role in driving global economic growth in the long run. The evolving technology is expected to boost productivity, create new revenue streams, and facilitate innovation.

Market research firm IDC, for instance, forecasts that in 2030, each dollar spent on AI-related services will generate $4.60 in value. And a report from the United Nations Trade and Development office suggests that the AI market could surge in value by 25x over the next decade, generating a whopping $4.8 trillion in revenue in 2033.

With that in mind, it won't be surprising to see organizations and governments investing far more money in AI-focused hardware and software to become more productive and efficient. That's why investors would do well to take a closer look at a couple of names that are playing central roles in the proliferation of AI.

Abstract representation of an integrated circuit with the term "AI" written on the processor.

Image Source: Getty Images

1. Broadcom

Broadcom (AVGO 1.63%) makes specialized application-specific integrated circuits (ASICs) and networking chips used in data centers, and demand for its processors has taken off thanks to the proliferation of AI systems. Several cloud computing giants have been deploying Broadcom's custom AI processors to lower the costs of AI training and inference, and to reduce their reliance on expensive graphics processing units (GPUs) from Nvidia.

ASICs are custom AI processors built to perform specific tasks, allowing them to deliver more computing power with lower energy consumption when compared to general-purpose GPUs. The advantages of custom AI processors make them ideal for large-scale deployment in data centers and are precisely the reason why Broadcom's AI business is taking off.

Currently, three hyperscale cloud customers use Broadcom's custom accelerators in large AI data centers to develop next-generation models. The chipmaker is deeply engaged with these customers to develop even more advanced custom processors and networking chips to support their product roadmaps for the next three years.

In Broadcom's view, these three hyperscalers alone should create a serviceable addressable market (SAM) worth $60 billion to $90 billion for it by fiscal 2027. The company's AI revenue jumped by 77% year over year in the first quarter of its fiscal 2025 to $4.1 billion, so it is currently clocking a more than $16 billion annual run rate.

Broadcom, therefore, has terrific room for growth in the custom AI chip market over the next three years. But it's also worth noting that the company is engaged with another four hyperscalers that are looking to build and deploy custom AI accelerators. Broadcom is in the final stages of chip development for two of those customers, while the other two recently selected the chipmaker to build their own custom chips.

So, Broadcom's addressable opportunity in custom AI chips could be much larger in the long run than what the company recently projected. As a result, it could end up delivering much stronger revenue growth over the next three fiscal years than what analysts currently expect.

AVGO Revenue Estimates for Current Fiscal Year Chart

Data by YCharts.

The semiconductor giant could easily exceed that $82 billion revenue forecast in three years, once it is producing custom AI chips in large volumes for all seven of its customers. This probably explains why the company trades at an attractive price/earnings-to-growth ratio (PEG ratio) of 0.64 based on its projected earnings growth for the next five years, according to Yahoo! Finance.

The PEG ratio is a forward-looking valuation metric that takes into account a company's expected earnings growth; a positive reading of less than 1 is generally viewed as an indication that a stock is undervalued. Broadcom's PEG ratio is well below that mark. As such, investors should consider buying this AI stock right away, before it flies higher following the 26% gain it clocked over the past month.

2. Lam Research

Lam Research (LRCX 1.33%) manufactures semiconductor manufacturing equipment -- machines used by foundries and chipmakers to make chips for everything from smartphones to cars to computers to data centers. And the market it operates in is on track to expand nicely due to the growing demand for AI chips.

According to one estimate, global spending on semiconductor equipment could jump to $121 billion in 2025 and to $139 billion in 2026. Those estimates point toward a nice improvement from last year, when spending stood at $113 billion. However, don't be surprised if semiconductor equipment spending increases at an even faster pace based on recent updates from key chip companies involved in the manufacturing of AI equipment.

Foundry giant Taiwan Semiconductor Manufacturing, popularly known as TSMC, plans to have eight new chip fabrication plants under construction this year, as well as one advanced chip packaging facility. Memory specialist Micron Technology, meanwhile, expects to lay out $50 billion in capital expenses in the U.S. through 2030. As such, it is not surprising that Lam Research saw an impressive acceleration in its revenue and earnings growth.

In its current fiscal year, analysts expect its sales to increase by 22% to $18.2 billion and its earnings to increase by 32%. What's more, management is confident that it will achieve revenue in the $25 billion to $28 billion range by 2028, indicating that its top line could jump by around 50% over the next three fiscal years.

Assuming Lam hits the midpoint of its 2028 forecast range and that the stock maintains its current price-to-sales ratio of 6.4 at that time, its market cap would increase to around $170 billion. That would amount to a jump of around 65% in the space of three years. However, Lam today trades at a cheaper price-to-sales ratio than the U.S. technology sector's average of 7.4. So it won't be surprising to see this semiconductor stock delivering even stronger gains than that, as the market could put a higher valuation on it in light of its robust growth.